00:00
News Radio 700 WLW.
So in the midst of, I think it's clear, there's a revolution of sorts when it comes to technology and the idea of AI, whether it's writing stories for magazines and newspapers, online to creations of all sorts of things. And maybe in many cases, the idea that the machines could take us over. Elon Musk and other experts in that field have asked for a pause, a delay in any further development of such for like six months or more, as they try to figure out, I guess, what the future may hold in to maybe keep the machines from taking over. But something that maybe is a little bit more relatable, that maybe is a little bit more of a danger than the machines taking us out? Is people using this artificial intelligence to create things that they would not have been able to otherwise, and then maybe cashing in and who has ownership to this, and where are the ethics associated with it, and a guy who deals with patents. And an interesting piece talking about the creation of things with AI is John Rizvi. He's an attorney kinda to give us some time dealing with he's known as we call yourself there, John, welcome to 700. W. L. W was Sterling. You are like the patent guy, right?
01:06
Yeah, the Patent Professor. Yes.
01:08
The patent professor? Well, let's get some education. I think it makes sense. We're due for some education in this. Where is the line? And I guess part of this conversation is there's a woman who put out a book and allowed the art to be created by artificial intelligence. And then the question is, who owns the art? Who should get paid? And? And what is the allowable use of arts created by the machines? Right?
01:31
Yeah, absolutely. There might be a little bit of history that might help to really understand the context. I don't know if you remember, a few years ago, there's that monkey selfie case, as it's known, where a photographer had forgotten their cell phone, it was picked up by a monkey, it took a bunch of selfies that ended up being incredibly valuable, and went viral. And because it was his cell phone that was used, he filed for the copyright. And the copyright office said that, you know, based on where the law derives, constitutionally, an author has to be a human being, and the copyright was denied. So that's kind of the crux of the issue with computer generated art and computer generated content, whether it's text or, or graphics, is how much of the creativity is done by human beings, and how much of it is done. I don't know, wouldn't don't even know if I'd call it creativity, or how much of the work is done by the computer, I say that creativity because they're basically crawling the internet and a database of billions of documents that are already created. And that's the base material. So there's a human in this aspect who is just putting in a simple query into the program, and not doing the actual work. Now, the proponents who are claiming that there should be copyright protection for this are saying that the computer is no different than a paintbrush would be for a painter. It's just a tool. And the actual creativity is still with the human being. And that's, you know, that's I don't know what you think about that. But I certainly think there's a difference between a paintbrush, a painter using a paintbrush, versus the sophistication of some of these AI programs where you simply would put in what you want drawn. And voila, in two minutes, you have the whole art piece done.
03:33
Oh, absolutely. I agree with that. I mean, there's a difference between, you know, using lysing, that brush and letting your hand do the work, rather than it being someone else and you putting your name on it. But if you buy the computer, you buy the software, and it creates it as an architect still with using software, helping to design my home or my business or my yard and landscaping, or is that then the computers and then you have to say, well, who created the concept in the program that those people are using their talents to create something else with it. It is a tool, but it gets weird.
04:04
It gets weird because it's a tool way more powerful than anything we've seen. I mean, there's—some people would say—well, you know what, you know, Photoshop is a program, and people use Photoshop to create, to manipulate images all the time. And I think the difference is the creative control. You actually have the mouse in your hand, your keyboard, you're making direct edits, there's a lot more activity. You're not simply putting in a query like, you know, 'draw me a frog with a horse's head,' and then two minutes later, voila, it's there. There's a fundamental difference between a tool like a paintbrush. They also use the example of a chisel as well. This is way more sophisticated than a chisel, where every single hit is done with your hands with a hammer, and there's a lot more creativity involved here. That's the difference. Who is the actual author? And is it a human? Because if not, then a machine cannot own the right to a copyright.
05:11
John Rizvi, by the way, is the Patent Professor, talking about the issues of artificial intelligence, creating art or things, stuff, and who owns the right to what that creation is generally speaking with Sterling on a Wednesday night, 700 WLW. So as we look at this, what if, say, for instance, I buy a 3D printer, and they're building houses with 3D printers, they're doing a lot of things with 3D printers, and you can get programs that would then help design things and so forth. And there are elements that you may pick and choose from one place or another to go with it. I guess it gets deep as far as who has ownership. I mean, who owns the technology, that is the artificial intelligence, and is it then who's chasing money or a copyright for say, like the story about the girl who put that story out there and it created the artwork to go along with it that sort of was the crux of this initial conversation?
06:06
Yeah, so. So, she's kind of testing the limits of, you know. The original application was granted a copyright. And then the copyright office did a 180, they changed course and said, 'Well, wait a minute, there's no human authorship here.' And, therefore, there's no copyrights because in that case, there was almost no creative control on her part of the output. So now what she's doing is going through, and now it's going to be what's really going to test how much is enough. There's a new program called stable diffusion where you scan in your own drawings. So, the initial drawings are extremely rough hand drawings that you scan in. And then you put in some queries, some terms, and it manipulates your original hand drawings. This is really going to tell kind of if that's enough because now there is that element of creative control where you have a lot more input at the beginning. It's still an incredibly powerful tool. So, it's nowhere close to, you know, like a paintbrush or a chisel. But it's still a lot more creative than simply putting in the terms and asking you to draw like going back to my example of a frog with a horse's head, right? In this case with stable diffusion, you could do a rough sketch with, you know, it could look like a kindergartener's crayon drawing of a frog with a horse's head, scan it into this program, and then use a text query that explains what you're trying to draw. The artificial intelligence would look at your hand-drawn sketch, look at the query, and based on those, would create a final drawing. And the hope is that this would then be enough creative control on the part of the human to provide authorship for the creator, and therefore, copyrights will apply.
08:17
Man, I can see all kinds of ways this can be argued and manipulated. John Rizvi is the Patent Professor, talking about, well, who can own art created by or anything created by artificial intelligence with Sterling on the Big One. So, say, for instance, what's the classic story like the guy who came up with carbon paper or whatever it was, or even Post-it notes, I think? They come up with it, they may have gotten a small bonus, it was a part of their job for a company that they work for, then the company owns what they created. And then for whatever period of time has ownership of that, correct?
08:48
That's correct. If you're in—- a lot of that depends on the employment agreement. But most employment agreements do include a clause that any intellectual property you produce during the course of your employment belongs to the employer.
09:03
And as far as say, for instance, right now, if you're looking to write something and go into, like, the open chat API, or whatever, and give them ideas, come up with questions, come up with writing. I mean, at this point, I have friends who are educators, and they have to basically check their students' work to see if it's been heisted from someplace. Now they have software to search for something that could have been stolen elsewhere, not just reworded works and thoughts to something that they're regurgitating. And there are so many levels and layers to this. It gets very complicated. I mean, it's really hard to fathom this Christina Cashton OVA, who is the graphic novelist behind this in the ZX Raya of the dawn, which is in the midst of this conversation about the art and the root words that she owns the words to what she wrote in those stories. But the creative portion and the art, which was artificially created, then is that open source now like open licensing for like art or photos that's on the internet that people for us? And if you're not careful, if you post it on social media or otherwise, someone can come after you. I mean, that's an area where, you know, somebody like you looks to get some billable hours, right?
10:10
Yeah. Well, and oh my god, there's just so much, so many different angles to this. One is whether Christina, for example, can get copyright protection herself. But another completely different issue is all of the existing material that exists that's being utilized to create this. There's a class action claiming infringement, and there's got to be compensation for the original artists because the computer doesn't have its own creativity. It's simply compiling from existing material that's out there. One big distinction in Casanova's case is that the copyright office, when they denied the copyright, one big factor was that the software program was called Mid-Journey. It does not allow any editing after the computer spits back the image. So there's no human interaction at all. That's a potential weakness in the copyright office's stance denying copyrights for AI-generated material because clearly these companies are learning from this case. They're following it, and their later programs allow for editing after the computer generates the content. So for Chat GPT, for example, if you put in terms that say give me a 150-word essay about photosynthesis and how it works, it's going to generate content, it's going to give that to you. One way to become the author is to take that raw material, and if you wanted it, reword it. And it doesn't take a lot before you've edited it, and now you've got the human control, and you can file for copyright. So that's one of the big drawbacks in Mid-Journey, which prevented the copyright office from finding any authorship, is that there's no editing permitted after the final product.
12:31
This is all crazy talk. You know, I mean, to a certain extent, it seems so out there, but it is. So now at this point, law is behind technology and science. It has always been that way. But it seems like it's really behind in some ways. I gotta ask because I mean, I know you put out books like Escaping the Gray, Thinking Grow Rich for Investors, you got The Pat Professor dot com, you were one of the original sharks, as I understand, on the Shark Tank. I mean, you've been around, you know, issues of patenting and so forth. We have problems enforcing that with China, nation-states. Other than that, let alone courts and other countries we have agreements with, where they don't follow along with our trade, trademarks, and copyrights and infringe upon those, not even for knockoff goods, just for other things in their markets as well. That, I mean, just with Russia recently, with all the stuff that went down with Ukraine, as far as McDonald's and so on selling more or less the same product, but yanking the Golden Arches off in a lot of weird stuff like that. This is the same type of concept, it's just a machine creating all of it. Right?
13:30
It is, and it just brings, you know, it's funny that to go back to the original constitutional basis for intellectual property for protecting the rights of authors. And it's going back to, you know, what did the founders envision by authors? They clearly did not envision an author to be a computer. They didn't, as the monkey selfie case kind of shows us, they didn't envision an author to be a monkey either. An author is a human being, and there has to be creative control exerted by a human in creating content. The question just becomes one of degrees, like at which point the tool becomes so incredibly powerful that it is doing a lot of the creative work that otherwise a human would be. Like, this is like, you know, some of these new AI-based programs are like Photoshop on steroids because the human element input is minor, and most of the work, most of the creative control is in the part of the machine. Now, there's that loophole that I mentioned, if this AI software allows for the human—after the content is fully created—to go back and make edits, now you're out. Now the copyright office is in a tough position, like, who's to determine that the editing is not enough creative control?
15:09
Well, yeah, look, I can pay an editor to help me fix a paper I've written, I'm still going to turn the paper in, the person who edited it gets a couple of bucks, right? So what's the difference? If it's technology doing the editing and me doing the creation that's being edited, right?
15:23
Right, well, the big difference would be is now if you edit it yourself, you own the copyright to it, and now they can't use they can't reject it based on there not being enough creative control by a human being because the program's allowed the human to step in once it's done and make changes. Then they can still get the copyright, and we all know it's a lot easier, you know, like the it's always easier being a critic than it is to create the original work. So to come back after the computer has done it and go back and make some minor inconsequential changes, that's easy. And if that's all it takes to entitle someone to a copyright, then that's a really low bar for them to meet in order to get entitled to protection.
16:13
Yeah, I mean, at this point, there's so much work that in the creative realm that is somewhat derivative anyway, it gets deep, and people fight those issues for trademark, copyright, and servicemark. And etc., to which obviously you know a lot about, the PatentProfessor.com, the web page. John Rizvi, thank you for making time. I don't know if I'm in a better place of figuring out what's my work, and what somebody else's, but it gets deep. It will be interesting to see how it plays out right now before the machines take over. And we all have a lot of free time on our hands.
Exactly.
Take care of yourself. I appreciate your time. Hope to have you back again soon. The PatentProfessor.com John Rizvi with Sterling on the Big One.